Hugging Face's logo

Langboat
/
bloom-389m-zh

Text Generation Transformers PyTorch Chinese bloom Inference Endpoints text-generation-inference
Model card Files Community
3

用Deepspeed读取模型进行微调，会把内存耗尽，报错 Step 1 exited with non-zero status 247 exits with return code = -9

#3
by x-lin - opened
Discussion
x-lin

使用未裁剪前的模型1b1不会出现这个问题

Sign up or log in to comment